NIK links inflammation to hepatic steatosis by suppressing PPARα in alcoholic liver

disease

Yaru Li1*, Mingming Chen2*, Yu Zhou1, Chuanfeng Tang1, Wen Zhang1, Ying Zhong1,

Yadong Chen3, Hong Zhou4, Liang Sheng1,5,6§

1Department of Pharmacology, School of Basic Medical Science, Nanjing Medical University,

Nanjing, Jiangsu 211166, China; 2Jiangsu Key Laboratory of Drug Screening, China

Pharmaceutical University, Nanjing, Jiangsu 210009, China; 3Laboratory of Molecular Design

and Drug Discovery, School of Science, China Pharmaceutical University, Nanjing, Jiangsu

211198, China; 4Department of Immunology, Nanjing Medical University, Nanjing, Jiangsu

211166, China; 5Key Laboratory of Rare Metabolic Diseases, Nanjing Medical University,

Nanjing, Jiangsu 211166, China; 6Department of Rehabilitation Medicine, Jiangsu Province

People's Hospital and Nanjing Medical University First Affiliated Hospital, Nanjing, Jiangsu

210029, China

*Yaru Li and Mingming Chen contributed equally to this work.

§Corresponding Author: Liang Sheng, Ph.D.

Department of Pharmacology, School of Basic Medical Science, Nanjing Medical University,

101 Longmian Ave, Nanjing, Jiangsu 211166, China

Tel.: (86)-13913007736; Fax: (86)-25-83237937

E-mail: lgsheng@njmu.edu.cn

Abstract

Background: Inflammation and steatosis are the main pathological features of alcoholic liver

disease (ALD), in which, inflammation is one of the critical drivers for the initiation and

development of alcoholic steatosis. NIK, an inflammatory pathway component activated by

inflammatory cytokines, was suspected to link inflammation to hepatic steatosis during ALD.

However, the underlying pathogenesis is not well-elucidated.

Methods: Alcoholic steatosis was induced in mice by chronic-plus-binge ethanol feeding.

Both the loss- and gain-of-function experiments by the hepatocyte-specific deletion,

pharmacological inhibition and adenoviral transfection of NIK were utilized to elucidate the

role of NIK in alcoholic steatosis. Rate of fatty acid oxidation was assessed in vivo and in

vitro. PPARα agonists or antagonists of MEK1/2 and ERK1/2 were used to identify the NIK-

induced regulation of PPARα, MEK1/2, and ERK1/2. The potential interactions between

NIK, MEK1/2, ERK1/2 and PPARα and the phosphorylation of PPARα were clarified by

immunoprecipitation, immunoblotting and far-western blotting analysis.

Results: Hepatocyte-specific deletion of NIK protected mice from alcoholic steatosis by

sustaining hepatic fatty acid oxidation. Moreover, overexpression of NIK contributed to

hepatic lipid accumulation with disrupted fatty acid oxidation. The pathological effect of NIK

in ALD may be attributed to the suppression of PPARα, the main controller of fatty acid

oxidation in the liver, because PPARα agonists reversed NIK-mediated hepatic steatosis and

malfunction of fatty acid oxidation. Mechanistically, NIK recruited MEK1/2 and ERK1/2 to

form a complex that catalyzed the inhibitory phosphorylation of PPARα. Importantly,

pharmacological intervention against NIK significantly attenuated alcoholic steatosis in

ethanol-fed mice.

Conclusions: NIK targeting PPARα via MEK1/2 and ERK1/2 disrupts hepatic fatty acid

oxidation and exhibits high value in ALD therapy.

Keywords: carnitine palmitoyl transferase 1α; mitogen-activated protein kinase/extracellular

signal-regulated kinase kinase; extracellular signal-regulated kinase; proinflammatory

cytokine.

Graphcial abstract

Ethanol expose activates hepatic NIK which recruits MEK1/2 and ERK1/2 to induce

inhibitory phosphorylation of PPARα.

Introduction

Ethanol intake is harmful at any dose and its risks rise with increasing levels of consumption

[1]. Excessive drinking causes alcoholic liver disease (ALD), which covers a spectrum of

pathological states encompassing alcoholic steatosis, alcoholic hepatitis, fibrosis, and

cirrhosis [2]. Alcoholic steatosis, defined histologically as the deposition of fat in small or

large droplets in hepatocytes, is the initial phase of ALD [3]. Excess fat accumulation in lipid

droplets induces hepatocellular ballooning to hinder blood flow and microcirculation in

sinusoidal space and consequently raises oxidative injury and endoplasmic reticulum stress in

hepatocytes [4, 5]. The latter two pathological events will aggravate ALD. Relieving alcoholic

steatosis thus could be effective to prevent or delay the progression of fatal ALD.

Considerable evidence indicates that alcohol exposure weakens the intestinal barrier and

facilitates the influx of lipopolysaccharide (LPS) [6], which increases the generation of

reactive oxygen species [7]. LPS and reactive oxygen species further stimulate Kupffer cells

to secrete proinflammatory cytokines, such as tumor necrosis factor α (TNFα) [6] and

interleukin 1β (IL1β) [8], which exacerbate inflammation and push the progression of

alcoholic steatosis [9]. Alcoholic steatosis is largely induced by the disruption of hepatic fatty

acid oxidation [10]; this disruption usually results from the damage to the function of

peroxisome proliferator-activated receptor α (PPARα), a primary controller of fatty acid

oxidation [11, 12]. Fatty acid oxidation is critical to protect hepatic lipid homeostasis from

excessive influx of fatty acids caused by alcohol-induced adipocyte lipolysis [13]. However,

there is little information concerning how inflammation affects hepatic fatty acid oxidation in

the pathogenesis of alcoholic steatosis.

NF-κB-inducing kinase (NIK), a Map3k14-encoded serine/threonine kinase, is aberrantly

activated in livers of mice and patients with ALD [14, 15] due to cytokine and chemokine

stimulation [16, 17]. Locating at the upstream of noncanonical NF-κB pathway, NIK

phosphorylates I-κB kinase α, subsequently initiating the phosphorylation and proteolytic

cleavage of p100 (NF-κB2 precursor) to produce p52 (active NF-κB2 isoform) for the

transcriptional regulation of target genes [18]. NIK may be an opportunity to understand the

role of inflammation in regulating alcoholic steatosis.

The present study revealed that NIK pushed aberrant fat accumulation in liver by disrupting

fatty acid oxidation during ALD, because NIK recruited and activated mitogen-activated

protein kinase/extracellular signal-regulated kinase kinase1/2 (MEK1/2) and extracellular

signal-regulated kinase 1/2 (ERK1/2) to suppress the fatty acid oxidation controller, PPARα.

Therefore, NIK could be a therapeutic target to stop inflammation from promoting alcoholic

steatosis.

Methods

Animals

All mouse experiments were conducted on the basis of relevant institutional and national

guidelines. The experimental protocol (Protocol Number: 1704009-3) was approved by the

Animal Care and Use Committee of Nanjing Medical University. NIK flox/flox mouse

(NIKf/f) in C57BL/6 background, was a present from Professor Liangyou Rui (University of

Michigan, Ann Arbor, MI, USA). The mouse was generated by inserting two loxp sites into

intron 1 and intron 2 that flank exon 2-6 of Map3k14. To generate hepatocyte-specific NIK-

deficient mice (NIKΔhep), we utilized albumin-cre transgenic mice (Jackson Laboratory, Bar

Harbor, ME, USA) to cross with NIKf/f mice. Wild-type (WT) mice in C57BL/6 background

were purchased from the Animal Core Facility of Nanjing Medical University, Nanjing,

China. In a pathogen-free barrier facility with controlled temperature and illumination, mice

had ad libitum access to sterile water and standard food. Following a previous study, male

mice aged 10 weeks received chronic-plus-binge ethanol feeding [19]. In detail as shown in

Figure S1A, mice fed a Lieber-DeCarli ethanol diet (5% ethanol, Trophic Animal Feed High-

Tech Co., Ltd, Haian, Jiangsu, China) for 10 d and thereafter were given a single gavage of

ethanol (5 g/kg body weight) on the eleventh day. Fenofibrate was orally administered

starting on the third day of ethanol feeding at a dose of 20 mg/kg/day [20]. The NIK inhibitor

B022, synthesized in accordance with a previous report [21], was dissolved in corn oil and

intraperitoneally administrated starting on the third day of ethanol feeding at a dose of 25

mg/kg/day.

Blood sample analysis

Blood was collected following decapitation. Serum levels of 3-hydroxybutyrate was assayed

using a kit (Megazyme International Ireland, Bray, Ireland).

Cell culture and treatment

AML12 cells (SCSP-550) and HepG2 (SCSP-510) were purchased from the Cell Bank of the

Chinese Academy of Sciences (Shanghai, China) and cultured according to the previous

studies [22, 23]. AML12 cells subjected to transfection were treated by stimuli 24 h latter, and

then further cultured for another 24 h. Serum starvation started 5 h before harvest. HepG2

cells infected by adenovirus were further cultured for 24 h. Serum starvation started 5 h

before harvest.

Primary hepatocytes were isolated by collagenase digestion from adult mice (8-10 weeks)

according to a previously published protocol [15]. After being cultured in William’s medium

E (Sigma-Aldrich, Shanghai, China) containing 6% FBS for 16 h, hepatocytes were subjected

to treatment with stimuli or adenoviral infection. Cells were harvested for subsequent assays

after 24-hour culture.

Plasmids

We purchased p3XFlag-CMV7.1 from Sigma-Aldrich and pcDNA3.1(+) from Invitrogen

(Carlsbad, CA, USA). Dr. Dongping Wei (The First Hospital of Nanjing, Nanjing, Jiangsu,

China) provided pcDNA-HA3 and pcDNA-HA3-RXRα and Dr. Liangyou Rui (University of

Michigan Medical School) provided pRK5-NIK, pRK5-NIK(KA), and β-gal expression

vectors. We subcloned pRK5-NIK and pRK5-NIK(KA) into pcDNA-HA3, pcDNA3.1(+), or

pAdeno-TBG-MCS-3FLAG (OBiO Technology Corp., Ltd, Shanghai, China). Bruce

Spiegelman provided pSG5 PPARα (Addgene plasmid # 22751), which was subcloned into

p3XFLAG-CMV7.1 (Sigma-Aldrich) and pcDNA-HA3. Vectors expressing PPARα mutants,

including PPARα (S6, 12, 21A), PPARα (S73, 76, 77A) and PPARα (S6, 12, 21, 73, 76, 77A),

were generated using p3XFLAG-CMV7.1-PPARα by Shanghai Generay Biotech Co., Ltd.,

Shanghai, China. A series of vectors expressing truncated PPARα (Figure S8) were prepared

using pcDNA-HA3-PPARα. Bruce Spiegelman also provided pcDNA-f:PGC1 (Addgene

plasmid # 1026), which was subcloned into pcDNA-HA3. John Kyriakis provided pMT

ERK1 (Addgene plasmid # 12656), which was subcloned into pcDNA-HA3. Melanie Cobb

provided pCMV-myc-ERK2-MEK1_fusion (Addgene plasmid # 39194), from which an

ERK2 fragment was subcloned into pcDNA-HA3, pCMV-3-tag-4A-myc, or pAdeno-

MCMV-MCS-3FLAG (OBiO Technology Corp., Ltd.). PPRE X3-TK-Luc was provided by

Bruce Spiegelman (Addgene plasmid # 1015). Fragments encoding mouse MEK1 and MEK2

were generated via polymerase chain reaction (PCR) from mouse liver cDNA and then

inserted into pcDNA-HA3 or pCMV-3-tag-4A-myc. A HiSCript II 1st Strand cDNA

Synthesis kit (Vazyme Biotech Co. Ltd, Nanjing, Jiangsu, China) was used to synthesize

cDNA for cloning, and PCR was performed using Phanta Max Super-Fidelity DNA

polymerase (Vazyme). Inserted fragments were cloned into vectors using a CloneExpress II

One Step Cloning kit (Vazyme). The primers used for cloning are listed in Table S1.

Generation of adenoviruses

Fragments encoding NIK or NIK(KA) were synthesized via PCR and inserted downstream of

the albumin promoter in the pAdeno-TBG-MCS-3FLAG vector. The empty pAdeno-TBG-

MCS-3FLAG vector and that containing the target genes were sent to OBiO Technology

Corp., Ltd. for adenovirus packaging and purification.

Adenoviral infection

Male mice aged 10 weeks were injected with adenoviruses (2 × 1010 viral particles [vp] for

each mouse) through tail vein. At day 5 after infection, mice were decapitated after 18 h of

fasting. Adenoviruses, with concentrations of 4 × 108 vp/well in 12-well plates or 8 × 108

vp/well in 6-well plates, infected primary hepatocytes and HepG2 cells. Cells were further

cultured for 24 h before harvest.

Immunoblotting and immunoprecipitation

We followed our previous protocol [24] to perform sample preparation, immunoblotting and

immunoprecipitation. The FLAG-tagged proteins and HA-tagged proteins were

immunoprecipitated by anti-FLAG M2 affinity gel (Sigma-Aldrich) and Pierce anti-HA

agarose (Thermo Fisher Scientific, Shanghai, China), respectively. Cytosol and nuclear

proteins were extracted by a kit (BioVision Incorporated, Milpitas, CA, USA). Bands in

immunoblots were quantified using Image J software (National Institutes of Health, Bethesda,

MD, USA; [1.37c]). The information concerning the antibodies and beads used is summarized

in Table S2.

Reverse transcriptional quantitative PCR

According to our previous protocol [24], we performed total RNA extraction, reverse-

transcription and PCR. P0 (36B4) large ribosomal proteins and glyceraldehyde-3-phosphate

dehydrogenase (GADPH) were internal controls for mouse samples and HepG2 cells,

respectively. The primer pairs used for reverse transcriptional quantitative PCR in this study

are listed in Table S3.

Luciferase assay

AML12 cells were transfected with PPRE X3-TK-Luc (200 ng/well), β-gal expression vector

(200 ng/well), and pSG5 PPARα (200 ng/well) plus other expression vectors (200 ng/well) as

indicated using polyethylenimine (Sigma-Aldrich). Then, after growth for 24 h, cells received

treatment of WY14643 (5 μmol/L, MedChemExpress), trametinib (100 nmol/L,

MedChemExpress), SCH772984 (50 nmol/L, MedChemExpress), or IKK16 (1 μmol/L,

MedChemExpress) for another 24 h. Luciferase activity was determined with luciferase

reporter assay system (Promega, Madison, WI, USA). β-gal activity, as the control for

transfection efficiency, was measured by a kit (Mairybio Biological, Beijing, China). Results

were averaged over three biological replicates.

Histopathological analysis and liver triglyceride (TAG) assay

Staining by Hematoxylin and eosin (H&E) and Oil Red O (Sigma-Aldrich) were performed

referring to the work of Sarmistha et. al. [25] to demonstrate lipid accumulation in the liver.

Following our protocol [24], liver TAG was extracted by chloroform-methanol solution and

determined by enzymatic method with kit.

β-oxidation assays in hepatocytes

The β-oxidation rate was determined according to our previous study with minor

modifications [26]. Briefly, hepatocytes were incubated in serum-free William’s medium E

containing 100 μmol/L palmitate conjugated with bovine serum albumin and 2 μCi/mL [9,

10-3H] oleic acid (American Radiolabeled Chemicals, Inc., St. Louis, MO, USA) at 37 ℃ for

1 h. Then the culture media were transferred, mixed with perchloric acid (1.3 mol/L), and

subjected to high-speed centrifugation. Thereafter, the supernatants were collected,

neutralized by potassium hydroxide (2 mol/L) plus 3-(N-Morpholino) propanesulfonic acid

(0.6 mol/L), and poured into an anion-exchange column (prepared with Dowex 1×8 anion-

exchange resin, Aladdin, Shanghai, China) to get rid of ketone bodies. The radioactivity of 3H

(tritium oxide) in the effluent normalized to protein amounts in the cells was determined to

calculate β-oxidation rates.

Far-western blotting

Far-western blotting was performed following a previously published protocol [27]. The

proteins that potentially interact with NIK, including PPARα, MEK1, and ERK2, were

expressed with a HA tag in AML12 cells and purified by immunoprecipitation. These target

proteins were transferred to polyvinylidene fluoride membrane by immunoblotting,

subsequently probed by a purified GST-infused NIK protein (4 μg/mL, Promega, Madison,

WI, USA), and eventually visualized using HRP-conjugated GST tag monoclonal antibodies

(#HRP-66001; Proteintech, Wuhan, China).

Statistical analysis

Results are demonstrated as the means ± SEM. Two groups of data were compared by two-

tailed Student’s t-test. More than two groups of data were compared by one-way analysis of

variance. p <0.05 indicates statistical significance.

Results

NIK mediates liver steatosis initiated by chronic-plus-binge ethanol feeding in mice

To investigate the role of NIK in alcoholic steatosis, we adopted an ALD mouse model of

chronic-plus-binge ethanol feeding [19]. As demonstrated in Figure S1A, mice received

chronic ethanol feeding for 10 d and got an acute binge at day 11. In WT mice, ethanol

feeding significantly increased the hepatic TAG content (Figure 1A) as well as p52 protein

level (Figure 1B), a classic biomarker for NIK activation. Consistently, hepatic NIK protein

level was also elevated by ethanol feeding (Figure S1C, upper panel) but the NIK mRNA

level remained unchanged (Figure S1B). Furthermore, we found that both chronic ethanol

feeding and acute binge contributed to the hepatic NIK upregulation (Figure S1C, lower

panel). In hepatocyte-specific NIK-deficient mice (NIKΔhep), ethanol-induced TAG

accumulation (Figure 1C) and aberrant p52 level increase (Figure 1D) in the livers were

significantly reduced compared to those in WT control mice (NIKf/f). Overexpression of NIK

significantly upregulated liver TAG and p52 levels (Figure 1E–F). These data indicate that

NIK activity is associated with TAG levels in the liver and that NIK may be a driver of

alcoholic steatosis.

To identify the NIK stimuli in intrahepatic environment of ALD, we utilized hepatocytes to

test a series of factors under appropriate pathological concentrations. Judging by the protein

levels of p52, we found that NIK was activated by hydrogen peroxide, palmitate, LPS, TNFα,

and IL1β, but not by ethanol or its metabolites. Among these activators, LPS, TNFα, and IL1β

exhibited the strongest activities (Figure S1D). LPS may activate NIK via Toll-like receptor

pathway [28]. TNFα and IL1β could enhance the stability of NIK protein [16, 17]. That may be

why ethanol feeding elevated NIK protein level but had no effect on NIK mRNA level in liver.

NIK promotes alcoholic steatosis via inhibition of fatty acid oxidation in the liver

To evaluate the regulation of hepatic fatty acid oxidation by NIK during alcoholic steatosis,

we determined the serum level of β-hydroxybutyrate and the expression of a series of

oxidative genes (carnitine palmitoyl transferase 1α, CPT1α; medium-chain acyl-coenzyme A

dehydrogenase, MCAD; long-chain acyl-coenzyme A dehydrogenase, LCAD; acyl coenzyme

A oxidase 1, ACOX1) as indicators for hepatic fatty acid oxidation capacity. Ethanol

consumption reduced the serum β-hydroxybutyrate level (Figure 2A) and downregulated the

hepatic mRNA levels of CPT1α and ACOX1 (Figure 2B) in WT mice. NIKΔhep mice exhibited

higher serum β-hydroxybutyrate levels (Figure 2C) as well as higher hepatic mRNA levels of

CPT1α and ACOX1 (Figure 2D) than NIKf/f mice after ethanol feeding. Overexpression of

NIK in the liver reduced serum β-hydroxybutyrate levels (Figure 2E) and CPT1α, LCAD,

ACOX1 mRNA expression (Figure 2F). CPT1α, as a key enzyme in the mitochondrial

oxidation of fatty acids, received our additional attention. The protein level of CPT1α

changed in line with its mRNA level (Figure 2B, D and F), besides, the protein level of the

exogenous recombinant CPT1α was not affected by co-expressed NIK in AML12 cells

(Figure S1E). Hence, NIK should regulate CPT1α expression at the transcriptional instead of

the post-transcriptional level. Taken together, these results indicate that NIK-mediated

suppression of hepatic fatty acid oxidation should contribute to ethanol-induced TAG

accumulation in the liver.

NIK reduces fatty acid oxidation by suppressing hepatic PPARα

A luciferase system was utilized to evaluate the role of NIK in the regulation of the

transcriptional activity of PPARα, the primary controller of fatty acid oxidation in the liver

[29]. This luciferase system was successful in assessing PPARα activity, as the luciferase

activity significantly increased when PPARα is overexpressed. NIK significantly reduced

PPARα-driven luciferase activity, and WY14643, a high-performance PPARα selective

agonist, protected PPARα activity from NIK (Figure 3A). NIK also suppressed fatty acid

oxidation in hepatocytes, which was reversed by WY14643 (Figure 3B). In addition, the

lowered mRNA and protein levels of CPT1α in hepatocytes due to NIK overexpression were

reversed by WY14643 (Figure S2A). The NIK-induced suppression of fatty acid oxidation in

hepatocytes was not caused by cell death because NIK overexpression did not reduce cell

viability (Figure S2B). The regulation of PPARα by NIK was further assessed in mice treated

with fenofibrate, a PPARα agonist used clinically. Hepatic steatosis induced by chronic-plus-

binge ethanol feeding or NIK overexpression in the liver was significantly attenuated by

fenofibrate (Figure 3C–D). Serum levels of β-hydroxybutyrate reduced by ethanol feeding or

hepatic NIK overexpression were elevated by fenofibrate (Figure 3E–F), and correspondingly,

the reduction in the mRNA and protein levels of CPT1α in the liver were reversed by

fenofibrate (Figure S2C–D). These results suggest that PPARα is the regulatory node of NIK

for pushing the process of alcoholic steatosis.

NIK suppresses the transcriptional activity of PPARα by phosphorylation

To explore the regulatory mode of NIK on PPARα, we focused primarily on the

phosphorylation of PPARα, as NIK is a kinase and it is reported that phosphorylation of the

PPARα serine residues disrupts its transcriptional activity [30]. Here, ethanol consumption

increased the serine phosphorylation of PPARα in the liver of mice, which was attenuated by

hepatocyte-specific NIK deletion (Figure 4A). Meanwhile, NIK overexpression enhanced the

serine phosphorylation of PPARα in the liver and hepatocytes (Figure 4A–B). Small

ubiquitin-like modifier (SUMO)-ylation, another inhibitory modification of PPARα [31], was

not enhanced by NIK (Figure 4B). In addition, other regulatory modes were tested. Protein

levels, and nuclear translocation of PPARα [32] in the liver were not affected by ethanol

feeding, NIK deletion or NIK overexpression (Figure S3A). Surprisingly, the interactions

between PPARα with its nuclear receptor heterodimer (RXRα) or its transcriptional

coactivator (PGC1α) [32-34] were not reduced but enhanced by NIK overexpression (Figure

S3B–C), implying that NIK-induced suppression was not via disrupting the interaction of

PPARα with its nuclear receptor heterodimer or coactivator. NIK-activated canonical and

noncanonical NF-κBs in hepatocytes by increasing the protein levels of p52 and RelA in

nuclei (Figure S3D) theoretically may suppress PPARα transcriptional activity by interfering

with the binding of PPARα to DNA [35]. However, IKK16, the inhibitor of I-κB kinase α and

I-κB kinase β that completely blocked NF-κB activation, did not significantly reverse the

inhibition of PPARα activity or fatty acid oxidation by NIK (Figure S3E–F). This suggests

that, in hepatocytes, NIK–NF-κB axis provides negligible contribution to the suppression of

PPARα, and phosphorylation is the primary mode employed by NIK to regulate PPARα. To

identify the mechanism underlying the ability of NIK to phosphorylate PPARα, we confirmed

the interaction of NIK with PPARα by immunoprecipitation (Figure 4C). However, NIK did

not directly bind to PPARα, as indicated by far-western blotting analysis (Figure 4D). It

implies that NIK and PPARα coexist in one complex, but NIK-induced phosphorylation of

PPARα relies on other kinases.

NIK induces the phosphorylation of PPARα via MEK1/2 and ERK1/2

To identify the mediators implicated in PPARα phosphorylation by NIK, a series of currently

known kinases that catalyze inhibitory phosphorylation of PPARα were screened [36]. Among

them, ERK1/2 was verified to be the unique kinase activated by NIK via phosphorylation

(Figure 5A). To authenticate the way by which NIK phosphorylates ERK1/2, we utilized

inhibitors against Raf-1 (LY3009120), MEK1/2 (trametinib), and ERK1/2 (SCH772984) to

disrupt the Raf-1–MEK1/2–ERK1/2 pathway. As shown in Figure 5B, trametinib and

SCH772984, but not LY3009120, prevented ERK1/2 phosphorylation caused by NIK;

besides, LY3009120 did not reduce NIK-induced MEK1/2 phosphorylation. These data

indicated that MEK1/2 was a necessary pathway for NIK to induce ERK1/2 phosphorylation,

and NIK phosphorylated MEK1/2 without involvement of Raf-1. Consistently, in the liver,

the phosphorylation levels of MEK1/2 and ERK1/2 were elevated during NIK activation in

the case of ethanol feeding and NIK overexpression, but were attenuated when NIK was

deleted (Figure S4). To elucidate the way of NIK to interact MEK1/2 and ERK1/2,

coimmunoprecipitation and far-western blotting were used. As shown by

coimmunoprecipitation (Figure S5A–B), NIK coexisted with MEK1/2 and ERK1/2 in one

complex. Based on the structural and functional similarities between MEK1 and MEK2 as

well as between ERK1 and ERK2, MEK1 and ERK2 were selected for far-western blotting

assay as the representatives of MEK1/2 and ERK1/2, respectively. The data indicated that

NIK directly bound MEK1 and ERK2 (Figure S5C). It suggests that NIK recruits MEK1/2

and ERK1/2 in a direct combination to form a ternary complex.

Besides the interaction with NIK, MEK1/2 and ERK1/2 also bound PPARα (Figure S6). To

confirm the mediating effect of ERK1/2 and MEK1/2 on NIK-induced phosphorylation and

suppression of PPARα, the inhibitors of MEK1/2 (trametinib) and ERK1/2 (SCH772984)

were used. Trametinib and SCH772984 attenuated PPARα phosphorylation (Figure 5C) and

revised the PPARα activity (Figure 5D) and fatty acid oxidation (Figure 5E) suppressed by

NIK. The ERK1/2-targeted phosphorylation sites in PPARα are serine residues, including S6,

S12, S21, S73, S76 and S77 [30], and PPARα mutants, such as PPARα (S6, 12, 21A), PPARα

(S73, 76, 77A) and PPARα (S6, 12, 21, 73, 76, 77A) were prepared by mutating related serine

residues into alanine residues. That NIK induced the phosphorylation of those serine residues

were verified, as the PPARα mutants, PPARα (S6, 12, 21A) and PPARα (S6, 12, 21, 73, 76,

77A), had significantly lower phosphorylation levels compared to the wildtype one under

NIK coexpression (Figure S7). PPARα (S6, 12, 21A) but not PPARα (S73, S76, S77A) was

resistant to NIK-mediated suppression (Figure 5F), suggesting that S6, S12 and S21 are

responsible for NIK's regulation over PPARα. Besides those ERK1/2-targeted serine residues,

NIK actually induces phosphorylation on other sites of PPARα, since PPARα with all the

ERK1/2-targeted sites mutated could still be phosphorylated by NIK (Figure S7).

PPARα is consist of four functional domains, including A/B, C, D, and E/F [37], and the C-

terminal of D region plus E/F region contain 12 helices (H1-H12) [38]. The motifs, so-called

D-box and T-box located in C region and D region, respectively, are involved in

heterodimerization [39]. H1-H2 in D region plays a role in the interaction with corepressors

[40]. The E/F region also has motifs involved in heterodimerization and interaction with

coactivator /corepressor, including H3-H5, H7-H9, H10-H11 (overlapping leucine-zipper

region), H12 (overlapping LLXXLL-binding pocket) [38, 39]. To identify the functional

motifs of PPARα to interact the NIK–MEK1/2–ERK1/2 ternary complex, we constructed a

series of vectors expressing truncated PPARα lacking A/B domain, D/T-box, H1-H2 region,

H3-H5 region, H7-H12 region, H10-H12 region, or H12 region (Figure S8). The

Coimmunoprecipitation indicated that PPARαΔA/B, PPARαΔH3-H5, PPARαΔH7-H12,

PPARαΔH10-H12 showed lower affinity for NIK compared to PPARαWT, while, the affinity

of PPARαΔD/T, PPARαΔH1-H2, and PPARαΔH12 for NIK did not decrease (Figure 5G–H).

Besides, PPARαΔH7-H12 and PPARαΔH10-H12 showed a similar degree of decline in

affinity for NIK (reduced by around 30%). These results suggest that A/B domain, H3-H5

region, and H10-H11 region are involved in the interaction of PPARα and the NIK-recruited

complex.

To determine whether NIK regulates PPARα in an analogous manner in human and mouse

hepatocytes, we utilized a hepatocyte cell line from human, HepG2. Similar to mouse primary

hepatocytes and liver, NIK overexpression reduced the rate of fatty acid oxidation and the

mRNA levels of related genes, and enhanced the phosphorylation levels of MEK1/2, ERK1/2

and PPARα (Figure S9). Slightly differently, NIK overexpression reduced the protein levels of

MEK1/2 and ERK1/2, which were not observed in mouse hepatocyte or liver (Figure S4 and

S9C). These data suggest that NIK also inhibits fatty acid oxidation and PPARα through

NIK–MEK1/2–ERK1/2 pathway in human liver cells.

NIK phosphorylates PPARα in kinase activity-dependent and -independent manners

To further evaluate the role of NIK in PPARα phosphorylation, the kinase deficient mutant of

NIK, NIK(KA), was compared with the WT one. Interestingly, NIK(KA), under the similar

expression level with NIK, did not lose but retain the ability to induce the phosphorylation of

MEK1/2, ERK1/2, and PPARα at a reduced level (Figure 6A–B). Therefore, NIK(KA) kept

some of the inhibitory actions of NIK, and exhibited a weaker suppression of PPARα activity

and fatty acid oxidation (Figure 6C–D). These results imply that NIK phosphorylates

MEK1/2, ERK1/2, and PPARα and inhibits hepatocyte fatty acid oxidation in kinase activity-

dependent and -independent manners. As NIK directly bound MEK1 and ERK2 rather than

PPARα in the complex (Figure 4C–D, Figure S5 and Figure S6), the regulatory modes

dependent or independent of NIK kinase activity may be largely due to NIK’s regulation of

MEK1/2 and ERK1/2. To test this hypothesis, the interaction between MEK1 and ERK2 was

assessed under overexpression of NIK or NIK(KA). NIK enhanced the interaction of MEK1

with ERK2, which was sustained by NIK(KA) (Figure 6E). It suggests that besides directly

phosphorylating MEK1/2 or ERK1/2, NIK may recruit MEK1/2 and ERK1/2 to enhance their

interaction and thus facilitate the phosphorylation of ERK1/2 by MEK1/2.

Pharmacological intervention against NIK attenuates alcoholic steatosis

To assess the therapeutic value of NIK for ALD treatment, we used B022, a chemical NIK

inhibitor, to treat mice fed with chronic-plus-binge ethanol diet. B022 attenuated ethanol-

induced hepatic steatosis (Figure 7A), elevated β-hydroxybutyrate level in the serum (Figure

7B), and enhanced CPT1α mRNA and protein levels in the liver (Figure 7C). B022 also

relieved NIK activation, blocked MEK1/2–ERK1/2 pathway and protected PPARα activity in

the liver as shown by the reduced nuclear levels of p52 as well as the decreased

phosphorylation levels of MEK1/2, ERK1/2, and PPARα (Figure 7C–D). Therefore, B022

successfully protected the hepatic capacity of fatty acid oxidation from ethanol feeding and

attenuated alcoholic steatosis by repressing the NIK–MEK1/2–ERK1/2 pathway.

Discussion

In the present study, we have presented evidence that NIK is responsible for liver steatosis

induced by chronic-plus-binge ethanol feeding in mice. Deletion of NIK in hepatocytes

attenuated liver steatosis after ethanol consumption by protecting the hepatic capacity of fatty

acid oxidation. PPARα, the primary transcriptional controller of fatty acid oxidation, is a

regulatory node of NIK, as PPARα agonists reverse NIK-mediated hepatic steatosis and

malfunction of fatty acid oxidation. NIK recruits MEK1/2 and ERK1/2 to form a complex

that induces the inhibitory phosphorylation of PPARα (Figure 7F). Pharmacological

intervention of NIK resulted in a prominent therapeutic effect on alcoholic steatosis.

Therefore, NIK may be a valuable therapeutic target for treating ALD.

The influx of gut-derived LPS induced by ethanol consumption initiates inflammation during

ALD. LPS stimulates Kupffer cells to secrete proinflammatory cytokines and chemokines,

such as TNFα and IL1β [6, 8, 41]. TNFα, IL1β, and LPS exhibit much higher efficiency to

activate hepatocyte NIK than ethanol and its metabolites. Moreover, preventing gut-derived

LPS flux by depleting intestinal microflora with antibiotics significantly attenuated alcoholic

steatosis [6]. These prove that enterohepatic axis-derived inflammation, as the most essential

cause of NIK activation, plays a key role in the development of ALD. Because hepatocyte

NIK deletion attenuated alcoholic steatosis, NIK likely links inflammation to ethanol-induced

liver steatosis. The efficacy of the treatment of alcoholic steatosis by a NIK inhibitor, B022,

further confirms that NIK has potential as a therapeutic target for ALD. However, NIK

deficiency in hepatocytes failed to improve nonalcoholic steatosis induced by a high-fat diet

[42], despite that NIK is activated both in nonalcoholic and alcoholic steatosis [14, 15]. These

results are presumably attributed to the differences in pathogenesis between the two fatty liver

models. Excessive adipocyte lipolysis induced by ethanol exposure [13] leads to alcoholic

steatosis when fatty acid oxidation is disrupted by aberrantly activated NIK in hepatocytes,

whereas high-fat diet feeding induces chronic TAG deposits largely though the enhancement

of hepatic lipogenesis resulting from the synergy and complementarity of the NIK pathways

in multiple liver cell types besides hepatocytes [42]. Therefore, simply deleting NIK in

hepatocytes is enough to attenuate alcoholic steatosis by rescuing fatty acid oxidation but is

insufficient to reverse high-fat diet-induced excessive lipogenesis contributed by diverse liver

cell types. It is likely that NIK is the better therapeutic target for alcoholic steatosis than for

nonalcoholic steatosis.

PPARα, the main controller of fatty acid oxidation in the liver, is regulated by NIK during

ALD, considering that PPARα agonists reversed NIK-mediated suppression of PPARα

activity, reduction in fatty acid oxidation, and hepatic steatosis. After screening several

potential regulatory modes, we determine that serine phosphorylation contributes to the

suppression of PPARα by NIK, as PPARα with triple mutations at S6, S12 and S21 is resistant

to NIK-mediated suppression. However, there are conflicted reports for the activity regulation

of PPARα by the phosphorylation of those serine residues. Juge-Aubry, C. E., et al. believed

that the phosphorylation at S12 and S21 enhanced PPARα activity [43], but Barger PM, et al.

demonstrated that phosphorylation at S6, S12 and S21 suppressed PPARα activity [30]. Our

results were consistent to the latter. As PPARα activity was regulated by a series of coactivators

or corepressors [40, 44-46], we speculate that phosphorylation of S6, S12, S21 may change the

configuration of PPARα leading to the dissociation of some corepressor, which may facilitate

the entry of other coactivator or corepressor. Whether coactivator or corepressor is bound to

PPARα may depend on the cell state. That should be why phosphorylation at those residues

causes different regulation of PPARα activity [30, 43]. In case NIK is activated, NIK may

recruit some potent corepressor facilitating its interaction with PPARα and thus suppress

PPARα activity. Of course, these hypotheses require further confirmation.

NIK does not directly phosphorylate PPARα. It integrates MEK1/2, ERK1/2, and PPARα into

a complex, in which, NIK directly binds MEK1/2 and ERK1/2 but not PPARα, therefore, the

phosphorylation of PPARα by NIK should be conducted through MEK1/2 and ERK1/2.

ERK1/2 is the unique kinase phosphorylated and activated by NIK among those reported to

catalyze inhibitory phosphorylation of PPARα [30, 36]. MEK1/2, the upstream kinase of

ERK1/2 and substrate of NIK [47], has been confirmed necessary for NIK-induced ERK1/2

phosphorylation. Thus, the phosphorylation should be transmitted along the NIK–MEK1/2–

ERK1/2 pathway to PPARα. The A/B domain, H3-H5 region, and H10-H11 region in PPARα

were verified contributable to the interaction of PPARα and NIK–MEK1/2–ERK1/2 complex.

The A/B region contributing 70% transcriptional activity of PPARα contains an activation

function-1 domain, which forms an amphiphilic α-helix probably interacting with coactivator

or corepressor [48], and NIK-induced phosphorylation also occurs in A/B region. H3-H5

region plays a role in the heterodimerization with RXRα [49], suggesting that part of

interaction between PPARα and the NIK-recruited complex could involve RXRα. H10-H11

region contains leucine-zippers, which is a putative docking motif of ERK1/2 [39]. These

data imply that NIK-recruited complex anchors PPARα at multiple sites, and these sites are

not necessarily close to phosphorylation sites of PPARα.

Of note, the phosphorylation of MEK1/2 and ERK1/2 is not entirely driven by NIK kinase

activity, as this ability of NIK is partially retained by its kinase deficient mutant, NIK(KA).

We speculate that the integrating function of NIK may be also contributable, because NIK and

NIK(KA) have a similar ability to enhance the interaction between MEK1 and ERK2, which

may promote spatial proximity of MEK1 to ERK2 and thus facilitate the phosphorylation of

ERK2 by MEK1. In fact, MEK1/2 is also the substrate of ERK1/2, however, ERK1/2

catalyzes an inhibitory phosphorylation of MEK1/2 on threonine residues at positions 292 or

386 [50] instead of the activating phosphorylation on serine residues at positions 217 and 221

that were detected by the phospho-MEK1/2 antibody used in the present study. Hence, the

increase in activating phosphorylation of MEK1/2 could be conducted by some NIK(KA)-

recruited kinase but not ERK1/2. To identify this unknown kinase recruited by NIK is

conducive to further understand NIK function, and the corresponding regulatory mechanism

remains to be explored.

NIK is constitutively subjected to degradation mediated by a complex that includes cellular

inhibitor of apoptosis 1/2 and tumor necrosis factor receptor-associated factors 2/3 in a

ubiquitination/proteasome-dependent manner [16, 17, 51]. Cytokine stimulation blocks this

degradation process, leading to NIK stabilization and activation [16, 17]. As NIK

phosphorylates and suppresses PPARα in kinase activity-dependent and -independent

manners, accelerating NIK degradation is expected to be more effective than simply

inhibiting NIK activity in the treatment of ALD or other NIK–MEK1/2–ERK1/2 complex-

involved diseases. In addition to regulating lipid-metabolism genes, NIK seems to control a

complex program coordinating the expression and secretion of proinflammatory factors,

which sustain or amplify the inflammatory states of the liver. Specifically, NIK stimulates

hepatocytes to release proinflammatory factors that trigger immune cell activation via a

paracrine mechanism. Immune cell-generated proinflammatory factors further activate more

immune cells to exacerbate liver inflammation [10]. The liver inflammation may in turn

stimulate NIK in hepatocytes to deteriorate the disorder in lipid metabolism during ALD.

Thus, NIK suppression has the dual effects of anti-inflammation and anti-fat deposition in

liver, and these effects are both beneficial to revise alcoholic steatosis in ALD therapy.

In conclusion, the aberrant activation of NIK may represent the underlying pathogenesis of

alcoholic steatosis in mice. NIK plays a causal role in impeding fatty acid oxidation by

restraining PPARα activity in hepatocytes. We identified the NIK–MEK1/2–ERK1/2 pathway

as the main signaling route through which NIK regulates PPARα. Our study suggests that

disruption of NIK in hepatocytes could offer new therapeutics for the treatment of ALD.

Abbreviations

ACOX1: acyl coenzyme A oxidase 1; ALD: alcoholic liver disease; Ad: adenovirus; CPT1α:

carnitine palmitoyl transferase 1α; ERK: extracellular signal-regulated kinase; EtOH: ethanol;

GADPH: glyceraldehyde-3-phosphate dehydrogenase; HRP: horseradish peroxidase; IB:

immunoblot; IL-1β: interleukin-1 β; IP: immunoprecipitation; LCAD: long-chain acyl-

coenzyme A dehydrogenase; LPS: lipopolysaccharide; MCAD: medium-chain acyl-coenzyme

A dehydrogenase; MEK: mitogen-activated protein kinase/extracellular signal-regulated

kinase kinase; mut, mutated; NIK: NF-κB-inducing kinase; n.s.: no significance; PGC1α:

peroxisome proliferator-activated receptor gamma coactivator 1α; PPARα: peroxisome

proliferator-activated receptor α; PPREs: PPAR response elements; RelA: v-rel

reticuloendotheliosis viral oncogene homolog A; PCR: polymerase chain reaction; RXRα:

retinoid-X receptor α; SDS-PAGE: sodium dodecyl sulphate-polyacrylamide gel

electrophoresis; SEM: standard error of the mean; SUMO: small ubiquitin-like modifier;

TAG: triglyceride; TNFα: tumor necrosis factor α; WT: wild-type.

Acknowledgments

This project was supported by the National Natural Science Foundation of China (Grant No.

81770862), and the Jiangsu Provincial Innovation Team Program Foundation. We sincerely

thank Dr. Liangyou Rui (University of Michigan) for NIKΔhep mice and expressing vectors. We

thank Dr. Dongping Wei (First Hospital of Nanjing) for expressing vectors. We greatly

appreciate Dr. Wei Gao (Nanjing Medical University), Dr. Zhong Li (Nanjing Medical

University), Dr. Chang Liu (China Pharmaceutical University) and Dr. Zheng Chen (Harbin

Institute of Technology) for insightful discussion.

Disclosures

The authors declare that there are no conflicts of interest.

References

1. Griswold MG, Fullman N, Hawley C, Arian N, Zimsen SR, Tymeson HD, et al. Alcohol use and burden for 195 countries and territories, 1990–2016: a systematic analysis for the Global Burden of Disease Study. Lancet. 2018; 392: 1015-35.

2. Jürgen R, Samokhvalov AV, Shield KD. Global burden of alcoholic liver diseases. J Hepatol. 2013; 59: 160-8.

3. Gao B, Ramon B. Alcoholic liver disease: pathogenesis and new therapeutic targets. Gastroenterology. 2011; 141: 1572-85.

4. Rangwala F, Guy CD, Lu J, Suzuki A, Burchette JL, Abdelmalek MF, et al. Increased production of sonic hedgehog by ballooned hepatocytes. J Pathol 2011; 224: 401–10.

5. Ikura Y, Caldwell SH. Lipid droplet-associated proteins in alcoholic liver disease: a potential linkage with hepatocellular damage. Int J Clin Exp Pathol 2015;8(8):8699-708

6. Gustot T, Lemmers A, Moreno C, Nagy N, Quertinmont E, Nicaise C, et al. Differential liver sensitization to toll-like receptor pathways in mice with alcoholic fatty liver. Hepatology. 2006; 43: 989-1000.

7. Thakur V, Pritchard MT, McMullen MR, Wang Q, Nagy LE. Chronic ethanol feeding increases activation of NADPH oxidase by lipopolysaccharide in rat Kupffer cells: role of increased reactive oxygen in LPS-stimulated ERK1/2 activation and TNF-alpha production. J Leukoc Biol. 2006; 79: 1348-56.

8. Mathews S, Gao B. Therapeutic potential of interleukin 1 inhibitors in the treatment of alcoholic liver disease. Hepatology. 2013; 57: 2078-80.

9. Xu MJ, Zhou Z, Parker R, Gao B. Targeting inflammation for the treatment of alcoholic liver disease. Pharmacol Ther. 2017; 180: 77-89.

10. Arsene D, Farooq O, Bataller R. New therapeutic targets in alcoholic hepatitis. Hepatol Int. 2016; 10: 538-52.

11. Crabb DW, Galli A, Fischer M, You M. Molecular mechanisms of alcoholic fatty liver: role of peroxisome proliferator-activated receptor alpha. Alcohol 2004; 34: 35-8.

12. Galli A, Pinaire J, Fischer M, Dorris R, Crabb DW. The transcriptional and DNA binding activity of peroxisome proliferator-activated receptor alpha is inhibited by ethanol metabolism. A novel mechanism for the development of ethanol-induced fatty liver. J Biol Chem. 2001; 276: 68-75.

13. Zhao C, Liu Y, Xiao J, Liu L, Chen S, Mohammadi M, et al. FGF21 mediates alcohol-induced adipose tissue lipolysis by activation of systemic release of catecholamine in mice. J Lipid Res. 2015; 56: 1481-91.

14. Shen H, Sheng L, Chen Z, Jiang L, Su H, Yin L, et al. Mouse hepatocyte overexpression

of NF-κB-inducing kinase (NIK) triggers fatal macrophage-dependent liver injury and fibrosis. Hepatology. 2014; 60: 2065-76.

15. Sheng L, Zhou Y, Chen Z, Ren D, Cho KW, Jiang L, Shen H, et al. NF-kappaB-inducing kinase (NIK) promotes hyperglycemia and glucose intolerance in obesity by augmenting glucagon action. Nat Med. 2012; 18: 943-9.

16. Vallabhapurapu S, Matsuzawa A, Zhang WZ, Tseng PH, Keats JJ, Wang H, Vignali DAA, et al. Nonredundant and complementary functions of TRAF2 and TRAF3 in a ubiquitination cascade that activates NIK-dependent alternative NF-κB signaling. Nat Immunol. 2008; 9: 1364-70.

17. Zarnegar B, Wang Y, Mahoney D, Pw, Cheung H, He J, Shiba T, et al. Noncanonical NF-kappaB activation requires coordinated assembly of a regulatory complex of the adaptors cIAP1, cIAP2, TRAF2 and TRAF3 and the kinase NIK. Nat Immunol. 2008; 9: 1371-8.

18. Xiao G, Fong A, Sun SC. Induction of p100 processing by NF-kappaB-inducing kinase involves docking IKKalpha to p100 and IKKalpha-mediated phosphorylation. J Biol Chem. 2004; 279: 30099-105.

19. Bertola A, Mathews S, Ki SH, Wang H, Gao B. Mouse model of chronic and binge ethanol feeding (the NIAAA model). Nat Protoc. 2013; 8: 627-37.

20. Zhang D, Tong X, Nelson BB, Jin E, Sit J, Charney N, et al. The hepatic BMAL1/AKT/lipogenesis axis protects against alcoholic liver disease in mice via promoting PPARalpha pathway. Hepatology 2018; 68: 883-96.

21. Ren X, Li X, Jia L, Chen D, Hou H, Rui L, et al. A small-molecule inhibitor of NF-κB-inducing kinase (NIK) protects liver from toxin-induced inflammation, oxidative stress, and injury. Faseb J. 2017; 31: 711-8.

22. Wu H, Ploeger JM, Kamarajugadda S, Mashek DG, Mashek MT, Manivel JC, et al. Evidence for a novel regulatory interaction involving cyclin D1, lipid droplets, lipolysis, and cell cycle progression in hepatocytes. Hepatol Commun, 2019; 3: 406-22.

23. Bessa MJ, Costa C, Reinosa J, Pereira C, Fraqa S, Fernández J, et al. Moving into advanced nanomaterials. Toxicity of rutile TiO2 nanoparticles immobilized in nanokaolin nanocomposites on HepG2 cell line. Toxicology and applied pharmacology. 2017; 316: 114-22.

24. Tang C, Liu P, Zhou Y, Jiang B, Song Y, Sheng L. Sirt6 deletion in hepatocytes increases insulin sensitivity of female mice by enhancing ERα expression. J Cell Physiol. 2019. 234: 18615-25. 

25. Mukherjee S, Chellappa K, Moffitt A, Ndungu J, Dellinger RW, Davis JG, et al. Nicotinamide adenine dinucleotide biosynthesis promotes liver regeneration. Hepatology. 2017. 65: 616-30.

26. Sheng L, Cho KW, Zhou Y, Shen H, Rui L. Lipocalin 13 protein protects against hepatic steatosis by both inhibiting lipogenesis and stimulating fatty acid beta-oxidation. J Biol Chem. 2011; 286: 38128-35.

27. Takahara T, Inoue K, Arai Y, Kuwata K, Shibata H, Maki M. The calcium-binding protein ALG-2 regulates protein secretion and trafficking via interactions with MISSL and MAP1B proteins. J Biol Chem 2017; 292: 17057-72.

28. Jie Z, Yang JY, Gu M, Wang H, Xie X, Li Y, et al. NIK signaling axis regulates dendritic cell function in intestinal immunity and homeostasis. Nat Immunol. 2018. 19: 1224-35.

29. Issemann I, Green S. Activation of a member of the steroid hormone receptor superfamily by peroxisome proliferators. Nature. 1990; 347: 645-50.

30. Barger PM, Brandt JM, Leone TC, Weinheimer CJ, Kelly DP. Deactivation of peroxisome proliferator-activated receptor-alpha during cardiac hypertrophic growth. J Clin Invest. 2000; 105: 1723-30.

31. Wadosky KM, Willis MS. The story so far: post-translational regulation of peroxisome proliferator-activated receptors by ubiquitination and SUMOylation. Am J Physiol Heart Circ Physiol. 2012; 302: 515-26.

32. Longuet C, Sinclair EM, Maida A, Baggio LL, Maziarz M, Charron MJ, et al. The glucagon receptor is required for the adaptive metabolic response to fasting. Cell Metab. 2008; 8: 359-71.

33. Tuomas VLI, Ferdinand M, Chris O, Carsten C, Mikael PK. Molecular mechanism of allosteric communication in the human PPARalpha-RXRalpha heterodimer. Proteins. 2010; 78: 873-87.

34. Vega RB, Huss JM, Kelly DP. The coactivator PGC-1 cooperates with peroxisome proliferator-activated receptor alpha in transcriptional control of nuclear genes encoding mitochondrial fatty acid oxidation enzymes. Mol Cell Biol. 2000; 20: 1868-76.

35. Delerive P, De Bosscher K, Besnard S, Vanden Berghe W, Peters JM, Gonzalez FJ, et al. Peroxisome proliferator-activated receptor alpha negatively regulates the vascular inflammatory gene response by negative cross-talk with transcription factors NF-kappaB and AP-1. J Biol Chem. 1999; 274: 32048-54.

36. Burns KA, Heuvel JPV. Modulation of PPAR activity via phosphorylation. Biochim Biophys Acta 2007; 1771: 952-60.

37. Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schütz G, Umesono K, et al. The nuclear receptor superfamily: the second decade. Cell. 1995; 83: 835-9.

38. Venäläinen T, Molnár F, Oostenbrink C, Carlberg C, Peräkylä M. Molecular mechanism of allosteric communication in the human PPARalpha-RXRalpha heterodimer. Proteins. 2010; 78: 873-87.

39. Burgermeister E, Lanzendoerfer M, Scheuer W. Comparative analysis of docking motifs in MAP-kinases and nuclear receptors. J Biomol Struct Dyn. 2003; 20: 623-34.

40. Liu MH, Li J, Shen P, Husna B, Tai ES, Yong EL. A natural polymorphism in peroxisome proliferator-activated receptor-alpha hinge region attenuates transcription due to defective release of nuclear receptor corepressor from chromatin. Mol Endocrinol. 2008; 22: 1078-92.

41. Uesugi T, Froh M, Arteel GE, Bradford BU, Thurman RG. Toll-like receptor 4 is involved in the mechanism of early alcohol-induced liver injury in mice. Hepatology 2001; 34: 101-8.

42. Liu Y, Sheng L, Xiong Y, Shen H, Liu Y, Rui L. Liver NF-κ B-Inducing Kinase Promotes Liver Steatosis and Glucose Counterregulation in Male Mice with Obesity. Endocrinology 2017; 158: 1207-16.

43. Juge-Aubry CE, Hammar E, Siegrist-Kaiser C, Pernin A, Takeshita A, Chin WW, et al. Regulation of the transcriptional activity of the peroxisome proliferator-activated receptor α by phosphorylation of a ligand-independent trans-activating domain. J Biol Chem, 1999; 274: 10505-10.

44. Huang Q, Alvares K, Chu R, Bradfield CA, Reddy JK. Association of peroxisome proliferator-activated receptor and Hsp72. J Biol Chem. 1994; 269: 8493–7.

45. Jia Y, Qi C, Kashireddi P, Surapureddi S, Zhu YJ, Rao MS, et al. Transcription coactivator PBP, the peroxisome proliferator-activated receptor (PPAR)-binding protein, is required for PPARalpha-regulated gene expression in liver. J Biol Chem. 2004; 279: 24427–34.

46. Xu HE, Stanley TB, Montana VG, Lambert MH, Shearer BG, Cobb JE, et al. Structural basis for antagonist-mediated recruitment of nuclear co-repressors by PPARalpha. Nature. 2002. 415: 813-7.

47. Malinin NL, Boldin MP, Kovalenko AV, Wallach D. MAP3K-related kinase involved in NF-kappaB induction by TNF, CD95 and IL-1. Nature 1997; 385: 540-4.

48. Hi R, Osada S, Yumoto N, Osumi T. Characterization of the amino-terminal activation domain of peroxisome proliferator-activated receptor alpha. Importance of alpha-helical structure in the transactivating function. J Biol Chem. 1999; 274: 35152-8.

49. Gorla-Bajszczak A, Juge-Aubry C, Pernin A, Burger AG, Meier CA. Conserved amino acids in the ligand-binding and tau(i) domains of the peroxisome proliferator-activated receptor alpha are necessary for heterodimerization with RXR. Mol Cell Endocrinol. 1999; 147: 37-47.

50. Lake D, Corrêa SA, Müller J. Negative feedback regulation of the ERK1/2 MAPK pathway. Cell Mol Life Sci 2016; 73: 4397-413.

51. Liao G, Zhang M, Harhaj EW, Sun SC. Regulation of the NF-kappaB-inducing kinase by tumor necrosis factor receptor-associated factor 3-induced degradation. J Biol Chem. 2004; 279: 26243-50.

Figure legends

Figure 1. Chronic-plus-binge ethanol feeding induces hepatic steatosis of mice by

activating NIK in the liver. (A, C, E) Representative staining of hematoxylin and eosin

(H&E) and Oil Red O; liver triglyceride (TAG) levels. Bar = 200 μm. (B, D, F)

Representative immunoblots for FLAG-tagged NIK, p52, p85 and tubulin in the liver. (A, B)

Wild-type (WT) mice received a chronic-plus-binge ethanol (EtOH-fed; n = 7) or control diet

(Pair-fed; n = 7). (C, D) NIKf/f (n = 7) and NIKΔhep (n = 7) mice received chronic-plus-binge

ethanol diets. (E, F) WT mice were infected with adenoviruses expressing FLAG-tagged NIK

(Ad-NIK; n = 6) or control viruses (Ad-control; n = 6) for 5 d. Values are demonstrated as

means ± SEM. *P <0.05, for comparisons with the control.

Figure 2. Chronic-plus-binge ethanol feeding reduces serum levels of β-hydroxybutyrate

and downregulates CPT1α by activating NIK in the liver. (A, C, E) Serum levels of β-

hydroxybutyrate. (B, D, F) The mRNA levels of CPT1α, MCAD, LCAD, ACOX1, and

representative immunoblot of CPT1α in the liver. (A, B) WT mice received a chronic-plus-

binge ethanol (EtOH-fed; n = 7) or control diet (Pair-fed; n = 7). (C, D) NIKf/f (n = 7) and

NIKΔhep (n = 7) mice received chronic-plus-binge ethanol diets. (E, F) WT mice were infected

with adenoviruses expressing FLAG-tagged NIK (Ad-NIK; n = 6) or control viruses (Ad-

control; n = 6) for 5 d. Values are demonstrated as means ± SEM. *P <0.05, for comparisons

with the control.

Figure 3. NIK-induced hepatic steatosis and suppression of fatty acid oxidation are

reversed by an agonist of PPARα. (A) A luciferase assay assessing PPARα activity when

NIK is overexpressed upon the treatment of WY14643 (5 μmol/L) (n=3 for each group). (B)

Hepatocytes infected with adenoviruses expressing FLAG-tagged NIK (Ad-NIK; n = 4) or

control adenoviruses (Ad-Control; n = 4) were exposed to WY14643 (5 μmol/L), and fatty

acid oxidation rates were determined. (C, D) Representative staining of H&E and Oil Red O;

liver TAG levels. Bar = 200 μm. (E, F) Serum level of β-hydroxybutyrate. (C, E) WT mice

fed with a chronic-plus-binge ethanol diet were treated with or without 20 mg/kg/day

fenofibrates (n = 7 for treated and control groups). (D, F) WT mice infected with

adenoviruses expressing FLAG-tagged NIK were treated with or without 20 mg/kg/day

fenofibrates (n = 7 for treated and control groups). Values are demonstrated as means ± SEM.

*P <0.05, for comparisons with the control.

Figure 4. NIK induces the phosphorylation of PPARα. (A) WT mice were fed with a

chronic-plus-binge ethanol or the control diet (left panel). NIKf/f and NIKΔhep mice were

subjected to chronic-plus-binge ethanol feeding (middle panel). WT mice were infected with

adenoviruses expressing NIK or control adenoviruses (right panel). Liver extracts were

immunoprecipitated with an anti-PPARα antibody and immunoblotted with antibodies against

phosphoserine, PPARα, or tubulin. (B) Hepatocytes isolated from WT mice were infected

with adenoviruses expressing NIK (Ad-NIK) or control adenoviruses (Ad-Control). Cell

extracts were immunoprecipitated with an anti-PPARα antibody and immunoblotted with

antibodies against phosphoserine, small ubiquitin-like modifier proteins (SUMO), PPARα, or

tubulin. (C) FLAG-tagged PPARα and HA-tagged NIK were coexpressed in AML12 cells.

Cell extracts were immunoprecipitated with anti-FLAG M2 affinity gel or Pierce anti-HA

agarose and immunoblotted with antibodies against HA, FLAG, or tubulin. (D) Enriched HA-

tagged PPARα and MEK1 (positive control) were subjected to far-western blotting analysis

using glutathione s-transferase (GST)-infused NIK as a probe and immunoblotted with

antibodies against GST or HA.

Figure 5. NIK induces the phosphorylation of PPARα via the MEK1/2-ERK1/2 pathway.

(A) The extracts of AML12 cells, transfected with a vector expressing HA-tagged NIK or a

control vector, were blotted with antibodies as indicated. (B) Hepatocytes were infected with

NIK (Ad-NIK) or control (Ad-Control) adenoviruses and simultaneously treated with 50

nmol/L SCH772984, 100 nmol/L trametinib, 500 nmol/L LY3009120, or a vehicle. Cell

extracts were blotted using antibodies as indicated. (C) AML12 cells, expressing HA-tagged

NIK and FLAG-tagged PPARα, were treated with 50 nmol/L SCH772984 or 100 nmol/L

trametinib. Cell extracts were immunoprecipitated with an anti-FLAG M2 affinity gel and

immunoblotted with antibodies against phosphoserine, FLAG, HA, or tubulin. (D) Results of

luciferase assays assessing PPARα activity when NIK and PPARα is overexpression with or

without the treatment with SCH772984 (50 nmol/L) or trametinib (100 nmol/L; n = 3 for each

group). (E) Hepatocytes infected with adenoviruses expressing FLAG-tagged NIK (Ad-NIK)

with or without the treatment with SCH772984 (50 nmol/L) or trametinib (100 nmol/L; n = 4

for each group). The fatty acid oxidation rates were determined. (F) Results of luciferase

assays assessing PPARα activity under coexpression of NIK with PPARα(WT) or

PPARα(S6,12,21A) or PPARα(S73,76,77A) (n = 3 for each group). (G, H) Flag-tagged NIK

was coexpressed with HA-tagged PPARα truncations as indicated in AML12 cells. Cell

extracts were immunoprecipitated with an anti-FLAG M2 affinity gel and immunoblotted

with antibodies against FLAG, HA, or tubulin. Values are demonstrated as means ± SEM.

*P <0.05, for comparisons with the control.

Figure 6. NIK induces the phosphorylation of MEK1/2, ERK1/2, and PPARα in kinase

activity -dependent and -independent manners. (A) Hepatocytes isolated from WT mice

were infected with adenoviruses expressing FLAG-tagged NIK (Ad-NIK) or NIK(KA) (Ad-

NIK[KA]) or control adenoviruses (Ad-Control) for 24 h. Cell extracts were immunoblotted

with antibodies as indicated. (B) The extracts of AML12 cells expressing FLAG-tagged

PPARα with HA-tagged NIK or NIK(KA) were immunoprecipitated with an anti-FLAG M2

affinity gel and immunoblotted with antibodies against phosphoserine, FLAG, HA, and

tubulin. (C) Results of luciferase assays assessing PPARα activity when NIK or NIK(KA) is

overexpressed with PPARα (n = 3 for each group). (D) Hepatocytes infected with

adenoviruses expressing FLAG-tagged NIK (Ad-NIK) or NIK(KA) (Ad-NIK[KA]) or control

adenoviruses (Ad-Control) (n = 4 for each group). The fatty acid oxidation rates were

determined. (E) AML12 cells expressing Myc-tagged MEK1 and HA-tagged ERK2 with HA-

tagged NIK or NIK(KA). Cell extracts were immunoprecipitated with an anti-Myc antibody

and immunoblotted with antibodies against Myc, HA or tubulin. Values are demonstrated as

means ± SEM. *P <0.05, for comparisons with the control.

Figure 7. B022, a NIK inhibitor, protects against ethanol-induced hepatic steatosis in

mice. WT mice fed with a chronic-plus-binge ethanol diet were administered with B022 (25

mg/kg) intraperitoneally starting on the third day of ethanol feeding (n = 8 for each group).

(A) Representative staining of H&E and Oil Red O; liver TAG levels. Bar = 200 μm. (B)

Serum level of β-hydroxybutyrate. (C) Representative immunoblots of p52, lamin B1, p-

MEK1/2, MEK1/2, p-ERK1/2, ERK1/2, CPT1α, and p85 in the liver; the mRNA levels of

hepatic CPT1α. (D) Liver extracts were immunoprecipitated with an anti-PPARα antibody

and immunoblotted with antibodies against phosphoserine, PPARα, or tubulin. Values are

demonstrated as means ± SEM. *P <0.05, for comparisons with the control. (E) Proposed

model for NIK action during the pathogenesis of alcoholic steatosis. Ethanol consumption

activates hepatic NIK, whose activity is suppressed by B022. NIK induces the inhibitory

phosphorylation of PPARα by recruiting MEK1/2, ERK1/2, and PPARα, which prevent the

PPARα-mediated transcription of genes related to fatty acid oxidation.

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7